Search

Spectral and structure function methods are implemented to compute the dissipation rate e from broadband, diverging-beam acoustic Doppler current profiler (ADCP) data collected at four sites in a high-flow tidal channel. This paper shows that middepth estimates of ϵ obtained from spectral and second-order structure function (SF2) methods are both lognormally distributed with comparable means and variances. Speed bin-averaged e values agree to within 16%, depending on the site and tidal phase (ebh/flood). The close agreement between the two independent methods provides further support for the argument put forward by McMillan et al.: that is, that the factor-of-2 difference between shear probe and (spectral) ADCP estimates of e was likely caused by spatial differences in turbulence levels. The agreement between the spectral and both second- and third-order structure function methods also supports the use of C2 = 2.0 for the SF2 universal constant. Notably, however, the SF3 method was less robust for these data. Two additional aspects of the SF2 approach are examined in some detail: 1) the differences from upstream- and downstream-facing beams are shown to arise from the Reynolds stress and 2) the inability of the ADCP to resolve small-scale motions does not affect the estimates of ϵ but yields apparent Doppler noise levels that--counterintuitively--decrease with increasing flow speed and increasing dissipation rate. A modified SF2 method that accounts for the variance associated with the unresolved scales removes the flow speed dependence and yields noise level estimates that agree with the spectral values. [ABSTRACT FROM AUTHOR]

The ability to estimate the rate of dissipation ( ε) of turbulent kinetic energy at middepth in a high-speed tidal channel using broadband acoustic Doppler current profilers (ADCPs) is assessed by making comparisons to direct measurements of ε obtained using shear probes mounted on a streamlined underwater buoy. The investigation was carried out in Grand Passage, Nova Scotia, Canada, where the depth-averaged flow speed reached 2 m s−1 and the Reynolds number was 8 × 107. The speed bin-averaged dissipation rates estimated from the ADCP data agree with the shear probe data to within a factor of 2. Both the ADCP and the shear probe measurements indicate a linear dependence of ε on the cube of the flow speed during flood and much lower dissipation rates during ebb. The ebb-flood asymmetry and the small-scale intermittency in ε are also apparent in the lognormal distributions of the shear probe data. Possible sources of bias and error in the ε estimates are investigated, and the most likely causes of the discrepancy between the ADCP and shear probe estimates are the cross-channel separation of the instruments and the high degree of spatial variability that exists in the channel. [ABSTRACT FROM AUTHOR]

A pulse-to-pulse coherent acoustic Doppler profiler has been developed for high-resolution particle velocimetry in the ocean, in particular for remote measurements of suspended sediment flux and turbulence in the nearshore and continental shelf bottom boundary layer. Acoustic backscatter estimates of suspended particle concentration and velocity are determined simultaneously from the phase and amplitude of the backscattered signal over an O(1 m)-long profile with subcentimeter resolution, at an ensemble-averaged rate of O(25 Hz). To characterize the performance of the profiler as a remote turbulent flux sensor, laboratory experiments were carried out in a particle-laden high--Reynolds number round jet. The results include comparisons between the acoustic Doppler and standard laboratory sensors, and to previous experimental and theoretical results for turbulent jets. The observed mean radial particle flux profile is found to be consistent with the computed mean flux profile for a turbulent jet. The measured entrainment rate is within 10% of the accepted value for turbulent round jets. Energy spectra of the turbulent motions demonstrate the 25/3 slope characteristic of the inertial subrange. The kinetic energy spectral densities obtained with the Doppler profiler match observations with a Sontek acoustic Doppler velocimeter and are in qualitative agreement with hotfilm measurements within the jet. Space--time domain images of particle flux exhibit well-defined coherent structures. Cospectral analysis demonstrates that these larger structures dominate the particle flux. [ABSTRACT FROM AUTHOR]

This paper considers the problem of accurately measuring the sediment transport over bedforms where flow evolves continuously both in time and space. For this purpose, we have developed a pulse-to-pulse coherent Doppler sonar system designed in bistatic geometry with two fan-beam transmitters symmetrically positioned on each side of a multielement receive array. The system resolves 2D velocity components within a ±20° (~0.5m by ~0.5m) swath. The software-defined radio data acquisition and control system limited us at present to eight independent receiver channels, and consequently the azimuthal resolution of the systemis 48. As a preliminary test of the sonar system, the system operation was simulated using a model developed to predict coherent sonar performance. The uncertainties with respect to the prescribed values and mean measurements in the model results were confined to 0.35 and 0.23 cm s-1, respectively, in the presence of strong shear (~150 s-1) and 50 cm s-1 horizontal flow. An important thing is that the model allowed us to test and develop the signal processing algorithms necessary to invert the multibeam sonar data. Using sand of 0.4-mm median diameter, the laboratory trials were carried out in active sediment transport conditions over dunes with 2-m wavelength and ~0.90m s-1 unidirectional flow velocities. The results presented here focus mainly on 2D velocity field and indicate an average 4% deviation from the wake law and 8% from independent observations made with the wide-band multifrequency coherent Doppler profiler (MFDop) instrument under similar flow conditions. [ABSTRACT FROM AUTHOR]

Coherent Doppler sonar is a useful tool for noninvasive measurement of ocean currents, sediment transport, and turbulence in coastal environments. Various methods have been proposed to separately address two of its inherent limitations: velocity ambiguity and measurement noise. However, in energetic turbulent flows, both factors may be present simultaneously. The presence of measurement noise complicates velocity ambiguity resolution, and conversely velocity ambiguity presents a challenge for existing noise suppression methods. A velocity estimator based on maximum a posteriori (MAP) estimation has been developed to resolve velocity ambiguity and suppress measurement noise simultaneously rather than separately. The estimator optimally combines measurements from multiple acoustic carrier frequencies and multiple transducers. Data fusion is achieved using a probabilistic approach, whereby measurements are combined numerically to derive a velocity likelihood function. The MAP velocity estimator is evaluated using a high-fidelity coherent Doppler sonar simulation of oscillating flow and data from a towing tank grid turbulence experiment where both velocity ambiguity and backscatter decorrelation were present. Time series and spectra from MAP velocity estimation are compared to those obtained with conventional Doppler signal processing. In addition to robustly resolving velocity ambiguity, the MAP velocity estimator is shown to reduce high-frequency noise in turbulence spectra. [ABSTRACT FROM AUTHOR]

A new formula is derived for the asymptotic form of the magnitude of an autocorrelation coefficient for coherent Doppler sonar. The autocorrelation magnitude is shown to be a biased estimator in the limit of infinite ensemble length. Numerical simulation of a Gaussian random process is used to verify the asymptotic formula and to show that a bias persists for finite pulse-pair averages. Validity of the asymptotic formula is also confirmed using a high-fidelity coherent Doppler sonar simulation, and from sonar measurements in a towing tank. It is shown that the distribution of observed autocorrelation coefficients is well predicted by a Gaussian random process once the autocorrelation bias has been removed. [ABSTRACT FROM AUTHOR]